The Hydrated Electron • Explore Related Articles • Search Keywords John M
Total Page:16
File Type:pdf, Size:1020Kb
PC68CH21-Herbert ARI 4 April 2017 10:5 ANNUAL REVIEWS Further Click here to view this article's online features: • Download figures as PPT slides • Navigate linked references • Download citations The Hydrated Electron • Explore related articles • Search keywords John M. Herbert and Marc P. Coons Department of Chemistry and Biochemistry, The Ohio State University, Columbus, Ohio 43210; email: [email protected] Annu. Rev. Phys. Chem. 2017. 68:447–72 Keywords First published online as a Review in Advance on aqueous electron, water cluster anions, radiation chemistry, DNA damage March 27, 2017 The Annual Review of Physical Chemistry is online at Abstract physchem.annualreviews.org Existence of a hydrated electron as a byproduct of water radiolysis was estab- https://doi.org/10.1146/annurev-physchem- lished more than 50 years ago, yet this species continues to attract significant Access provided by Ohio State University Library on 05/03/17. For personal use only. Annu. Rev. Phys. Chem. 2017.68:447-472. Downloaded from www.annualreviews.org 052516-050816 attention due to its role in radiation chemistry, including DNA damage, and Copyright c 2017 by Annual Reviews. because questions persist regarding its detailed structure. This work provides All rights reserved an overview of what is known in regards to the structure and spectroscopy of − the hydrated electron, both in liquid water and in clusters (H2O)N , the latter of which provide model systems for how water networks accommodate an ex- cess electron. In clusters, the existence of both surface-bound and internally bound states of the excess electron has elicited much debate, whereas in bulk water there are questions regarding how best to understand the structure of the excess electron’s spin density. The energetics of the equilibrium species e −(aq) and its excited states, in bulk water and at the air/water interface, are also addressed. 447 PC68CH21-Herbert ARI 4 April 2017 10:5 1. INTRODUCTION 1.1. History The nature of the solvated electron is an old question. The first observation of what would eventually be understood, more than a century later, as a solvated electron dates to Humphry Davy, in whose laboratory notebooks from 1808 can be found a description of the “beautiful metallic appearance” and “fine blue colour” observed when potassium crystals are heated in the presence of ammonia vapor (1). Following the liquefaction of ammonia, the blue color of sodium/ammonia mixtures was noted by Weyl in 1864 (2), who attributed the blue color to formation of a chemical compound, NaNH3. This idea held sway for some time until convincing evidence against it was finally presented by Kraus in 1908 (3), in experiments originally intended to test the idea that electrons are the charge carriers in metallic conduction (4). The existence of dissolved ions as the charge carriers in electrolyte solutions had been established much earlier by Kohlrausch (5), but in Kraus’s view, “knowledge of the solid state of matter ... is far too limited to enable us to determine the nature of the processes which are specifically involved when electricity passes through a metal” (4, p. 1558), and he supposed that solutions of metals in nonconducting solvents might provide simpler systems on which to test theories of electrical conduction in metals. To this end, Kraus measured the electrical conductivity of solutions of alkali metals dissolved in liquid ammonia, and already in the first of these papers (in 1908), he proposed that the charge carriers were “electrons surrounded by an envelope of ammonia” (6, p. 1332), i.e., solvated elec- + trons, formed via the dissociation equilibrium Na Na + e −. This inference seems all the more profound when one considers that the nature of the electron as the charge carrier in cathode ray tubes had been established only about ten years earlier (7), and Kraus’s 1908 paper predates the publication of Millikan’s oil drop experiment (8). A few years after Kraus’s proposal, Gibson & Argo (9, 10) measured optical absorption spectra of solutions of alkali and alkaline earth metals dissolved in liquid ammonia and in organic amines. These solutions each exhibit a strong blue color, and the wavelength of maximum absorption in a given solvent is independent of the identity of the metal (10). Based on the classical theory of dispersion in metals, Gibson & Argo showed that the absorption cross sections could not be reconciled with the conductivity data under the assumption that only the undissociated metal was present in solution (9). With this, the notion of a solvated electron as a distinct chemical species was established. These experiments predate the development of the “new” quantum theory, but by 1946 the idea had been put forward that the optical spectra of alkali metal solutions in ammonia arises from s → p excitation of a particle in a quasi-spherical solvent void (11). This notion was later elaborated (12) and adapted for the aqueous electron (13) by Jortner and co-workers. − Access provided by Ohio State University Library on 05/03/17. For personal use only. Annu. Rev. Phys. Chem. 2017.68:447-472. Downloaded from www.annualreviews.org This review focuses on the specific case of the solvated electron in water, e (aq), for which detailed historical accounts of early experiments can be found elsewhere (14–16). Briefly, as early as 1952 it was suggested that such a species might be a byproduct of the radiolysis of aqueous solutions (17), but reaction of alkali metals with water does not produce any visible coloration, which is ultimately a consequence of the fact that the lifetime of e −(aq) in neutral water is ∼300 µs when generated in low concentrations and can be significantly shorter under other conditions (18). Aqueous electrons can be generated by pulsed radiolysis of aqueous solutions (18), as in the original 1962 measurement of the optical spectrum (19); by sonolysis of aqueous solutions (20); by two- photon laser excitation of liquid water; or else by photodetachment of a suitable electron donor, − − e.g., by accessing charge-transfer-to-solvent excited states of I (aq)orCN (aq) (21–23). In the 50+ years since it was first detected experimentally, e −(aq) has come to be recognized as one of the primary radicals formed upon radiolysis of water (14, 24, 25). 448 Herbert · Coons PC68CH21-Herbert ARI 4 April 2017 10:5 Theoretical attempts to understand the detailed structure of e −(aq) are nearly as old as the first measurement of the optical spectrum. Early models were continuum or semicontinuum in nature (13, 26–29) and assumed that the electron inhabits an excluded volume in the structure of − Cavity model: liquid water, consistent with experimental measurements of the partial molar volume of e (aq) (30, amodelinwhichthe 31). In the 1970s, analysis of electron spin resonance (ESR) spectra of electrons trapped in glassy aqueous electron alkaline water seemed to confirm this picture (32). The same picture would later emerge from creates and occupies a atomistic simulations using one-electron (pseudopotential) models (33, 34), and finally from first- quasi-spherical region principles calculations based on density functional theory (DFT) (35–37). As such, this cavity of excluded volume within the solvent model of e −(aq) has become the conventional paradigm, although not without occasional (and ongoing) controversy (38–42), as discussed in Section 4. Vacuum level: − reference energy of a In the 1980s, finite cluster analogues of the hydrated electron, (H2O)N , were created in a noninteracting molecular beam (43), and the possibility of mass-selecting these clusters and interrogating them electron removed from spectroscopically provides another avenue to understanding how water accommodates an extra the solvent electron. Interpretation of the cluster spectroscopy has proven controversial (44–46) and is dis- cussed in Section 2. Spectroscopic studies of e −(aq) in liquid water are discussed in Section 3. 1.2. Chemical Significance Although this review focuses mainly on structural, spectroscopic, and energetic considerations, the importance of e −(aq) as a potent reducing agent in aqueous chemistry cannot be overstated. Figure 1 illustrates some of the early-time events in water radiolysis (47), whereupon ionizing radiation generates three primary radicals: H•,HO•,ande −(aq). The latter thermalizes on a picosecond timescale, ∼3.5 eV below vacuum level, and its formation and depletion are readily monitored via an intense absorption at 720 nm that is ascribed to s → p excitation within the excluded volume of the cavity occupied by e −(aq). Although some reactions involving e −(aq) are diffusion-limited, many exhibit activation ener- gies of 1–8 kcal/mol (24), suggesting that the kinetics is controlled by the availability of a vacant orbital on the reacting partner species. However, even for the simplest reaction, + − • H (aq) + e (aq) → H (aq), 1. the molecular-level mechanism is not always clear. Whereas a long-held view is that e −(aq) always reacts via an electron-transfer mechanism (14), recent DFT simulations of reaction (1) in small water clusters suggest that it may occur via proton transfer into the e −(aq) cavity (48, 49). This requires significant (and simultaneous) rearrangement of both the proton and the electron hydra- tion shells, including considerable distortion of the unpaired electron that facilitates reaction but comes with an energetic penalty for desolvation of the two hydrophilic reactants. Perhaps sur- Access provided by Ohio State University Library on 05/03/17. For personal use only. Annu. Rev. Phys. Chem. 2017.68:447-472. Downloaded from www.annualreviews.org + − prisingly, Reaction 1 proceeds more slowly than H (aq) + HO (aq) → H2O, where one might anticipate similar disruption of the solvation shells around the two ions, and also more slowly than e −(aq) + HO•(aq) → HO−(aq).